Density-dependent dispersal in host-parasitoid assemblages

Most spatial population models assume constant rates of dispersal. However, in a given community, dispersal may not only depend on the density of conspecifics. i.e. density-dependent dispersal, but also on the density of other species, a phenomenon we term 'community-dependent dispersal'. We co-vary the densities of both the beetle host Callosobruchus chinensis and its parasitoid wasp. Anisopteromalus calandrae, in a laboratory study and record the proportions of each species that disperse within a two-hour period. The parasitoid in these systems exhibits community-dependent dispersal - dispersing more frequently when parasitoid density is high and larval host density is low. This supported our prediction that individuals should disperse according to competition for available resources. However, in this study the host's dispersal was independent of density. We suggest that this may be due to less intense selection acting on host dispersal strategies than in the parasitoid. We consider some possible consequences of community-dependent dispersal for a number of spatial population processes. A well-known host-parasitoid metapopulation model is expanded so that it includes a greater range of dispersal functions. When the model is parameterised with the parasitoid community-dependent dispersal function observed in the empirical study, similar population dynamics are obtained as when fixed-rate dispersal functions are applied, The importance of dispersal functions for invasions of both competitive and host-parasitoid systems is also considered. The model results demonstrate that understanding how individuals disperse in response to different species' population densities is important in determining the rate of spread of an invasion. We suggest that more empirical studies are needed to establish what determines dispersal rate and distance in a range of species, combined with theoretical studies investigating the role of the dispersal function in determining spatial population processes.